MXNet R Tutorials

These tutorials for the mxnet R package aim to teach both practical usage of neural networks in R as well as the basic internal building blocks of MXNet. Each tutorial is a self-contained Juypter Notebook which you can run and modify yourself; simply download all the Notebooks into a local folder by clicking here.

• Develop a Neural Network with MXNet in Five Minutes
• Handwritten Digits Image Classification
• Classify Images with a PreTrained Model
• Character-level Language Model using RNN
• LSTM Time Series Example
• NDArray: Vectorized Tensor Computations on CPUs and GPUs
• Symbol and Automatic Differentiation
• Callback Functions
• Custom Loss Function
• Custom Iterator Tutorial

Note: To run the tutorial Notebooks on your own machine, you must first install the mxnet R package plus Jupyter Notebook with the R kernel. Subsequently tell Juypter to use the R kernel after opening each tutorial notebook (after opening the notebook, click on: Kernel > Change Kernel> R).

More tutorials and examples are available in the GitHub repository. For more information about MXNet, check out the main website. The textbook Dive into Deep Learning is another good resource to learn about the mathematical concepts underlying neural networks and machine learning.